3.5 HW – Interactive Plots

Submission: Completed qmd file and rendered html file via Canvas.

Overview

These problems apply the idea from the notes ‘3.5 – Interactive Plots’, where we learned how to create interactive plots in R using plotly.

This assignment will visualize some important aspects of NFL players, including physical attributes, positional information and salary. We will be using a dataset of Madden 23 player info. A subset of the madden-23.csv data is shown below:

Full Name	Team	Position	college	awareness_rating	throwPower_rating	kickReturn_rating	leadBlock_rating	strength_rating	bCVision_rating	catchInTraffic_rating	playAction_rating	pursuit_rating	plyrAssetname	mediumRouteRunning_rating	catching_rating	acceleration_rating	spinMove_rating	height	finesseMoves_rating	spectacularCatch_rating	runBlock_rating	tackle_rating	injury_rating	zoneCoverage_rating	weight	plyrBirthdate	runningStyle_rating	deepRouteRunning_rating	firstName	yearsPro	totalSalary	trucking_rating	throwAccuracyShort_rating	position	jukeMove_rating	playRecognition_rating	shortRouteRunning_rating	status	lastName	jerseyNum	breakSack_rating	speed_rating	runBlockPower_rating	jumping_rating	toughness_rating	throwOnTheRun_rating	manCoverage_rating	stiffArm_rating	powerMoves_rating	iteration	release_rating	hitPower_rating	throwAccuracyMid_rating	kickAccuracy_rating	passBlockPower_rating	impactBlocking_rating	stamina_rating	carrying_rating	breakTackle_rating	plyrPortrait	kickPower_rating	plyrHandedness	throwUnderPressure_rating	team	signingBonus	passBlock_rating	changeOfDirection_rating	press_rating	throwAccuracyDeep_rating	archetype	blockShedding_rating	runBlockFinesse_rating	teamId	agility_rating	fullNameForSearch	overall_rating	passBlockFinesse_rating	age	primaryKey
Aaron Donald	Rams	RE	Pittsburgh	99	20	10	23	99	49	23	6	98	DonaldAaron_10852	9	42	90	25	73	97	47	45	93	98	29	280	########	Default Stride Loose	5	Aaron	8	41670000	32	6	RE	28	99	16	published	Donald	99	12	82	45	78	97	6	25	30	99	launch-ratings	12	89	6	21	45	93	88	48	17	6394	25	Right	15	Rams	45500000	45	68	21	6	DE_PowerRusher	97	45	24	86	Aaron Donald	99	45	31	10852
Davante Adams	Raiders	WR	Fresno State	99	45	45	25	63	94	99	6	44	AdamsDavante_10823	97	99	92	86	73	10	99	48	42	93	24	215	########	Default Stride Loose	95	Davante	8	97870000	63	15	WR	95	26	98	published	Adams	17	22	90	20	96	88	10	18	75	10	launch-ratings	99	15	10	30	13	45	97	76	85	6328	25	Right	20	Raiders	42350000	38	97	25	6	WR_DeepThreat	52	12	23	95	Davante Adams	99	17	29	10823
Myles Garrett	Browns	RE	Texas AM	98	55	10	17	96	39	21	10	97	GarrettMyles_12520	8	52	91	40	76	91	25	45	92	88	51	272	########	Long Stride Default	4	Myles	5	84270000	36	31	RE	36	94	15	published	Garrett	95	12	88	45	92	87	23	51	48	98	launch-ratings	31	84	30	20	45	87	88	55	34	8560	20	Right	16	Browns	65590000	45	71	44	27	DE_PowerRusher	90	45	5	85	Myles Garrett	99	45	26	12520
Trent Williams	49ers	LT	Oklahoma	99	12	10	98	98	10	15	6	33	WilliamsTrent_9800	5	20	74	10	77	10	51	98	39	87	14	320	########	Default Stride Awkward	5	Trent	12	97870000	10	6	LT	10	19	15	published	Williams	71	12	76	98	83	98	6	10	10	10	launch-ratings	15	60	6	10	95	99	88	40	7	3231	11	Right	14	49ers	40100000	93	58	10	6	OT_Power	32	99	15	73	Trent Williams	99	92	34	9800
Cooper Kupp	Rams	WR	Eastern Wash.	99	59	84	34	66	96	88	38	31	KuppCooper_12558	98	99	96	87	74	10	88	65	31	91	26	208	########	Default	94	Cooper	5	89850000	55	42	WR	94	22	99	published	Kupp	10	22	89	32	82	88	40	15	77	10	launch-ratings	90	26	36	20	39	54	99	76	86	8626	22	Right	29	Rams	32200000	46	94	10	31	WR_Slot	28	28	24	95	Cooper Kupp	98	34	29	12558
Jalen Ramsey	Rams	CB	Florida State	98	27	80	22	71	72	55	16	79	RamseyJalen_17538	24	80	95	69	73	41	70	34	75	88	99	208	########	Short Stride Default	12	Jalen	6	71000000	42	33	CB	78	96	36	published	Ramsey	5	34	92	33	94	90	24	98	57	35	launch-ratings	27	82	26	17	33	65	94	71	68	8185	18	Right	17	Rams	43600000	21	92	99	21	CB_MantoMan	50	34	24	91	Jalen Ramsey	98	34	27	17538

Assignment

The goal is to recreate some interactive plots with minimal guideance and some plots of your creation in order to understand the structure of plotly function calls. It is designed to be open ended so you can explore a variety of different plot types and interactive plot techniques.

Part a – Scatterplot

Start with a scatterplot visualizing the relationship between overall_rating and totalSalary. Your final plot must consider the following:

• Must be interactive.
• Must take into account overplotting, either by changing the plot type or utilizing additional aesthetics.
• Depending on your implementation of above, can you add additional mouse-over text to add to the plot?

# option 1: alpha blending with plotky
data_madden %>% 
  plot_ly(x = ~overall_rating,
          y = ~(totalSalary),
          text = ~`Full Name`) %>% 
  add_markers(alpha = 0.3)

# option 2: hexmap with ggplotly
p <- ggplot(data = data_madden,
            aes(x = overall_rating,
                y = totalSalary)) + 
  geom_hex() + 
  theme_bw()
ggplotly(p)

OPTIONAL CHALLENGE:

• Use slice_max() to look at the top 500 players according to overall_rating and recreate the following scatterplot / data table setup that can be used during EDA to identify outlying data points.

  • Try lasso-selecting a few points and seeing the results in the table!

shared_data <- data_madden %>% 
  slice_max(order_by = overall_rating,
            n = 500) %>% 
  highlight_key()

p <- shared_data %>% 
  plot_ly(x = ~overall_rating,
          y = ~totalSalary) %>% 
  add_markers(alpha = 0.3) %>%
  highlight(on = "plotly_selected")

bscols(p, datatable(shared_data, height = 500))

Part b – Dumbell chart

Recreate the following dumbell chart to the best of your ability. Features of this plot:

• Data needs to be pre-summarized to get the average totalSalary and signingBonus by Position. Calculate the number of players included in each average calculation as well.
• Sorted by increasing average totalSalary. HINT: This requires a mutate() statement, not an arrange() statement.
• For each position, the dark green plots the average totalSalary and light green adds the average signingBonus. Thus we can interpret the ending value as the average total money earned during the first year of the contract.
• Well formatted mouse-over interactivity features. HINT: Use html formatting and scales::comma().

Try to get your plot as close as you can to this plot (with a reasonable amount of effort :)

data_madden %>% 
  summarize(.by = Position,
            across(c(totalSalary, signingBonus), mean),
            n = n()) %>% 
  mutate(Position = fct_reorder(Position, totalSalary)) %>% 
  plot_ly() %>% 
  add_segments(x = ~totalSalary,
               xend = ~totalSalary + signingBonus,
               y = ~Position,
               yend = ~Position,
               color = I("grey")) %>% 
  add_markers(x = ~totalSalary,
              y = ~Position,
              text = ~paste0("Avg salary: $", scales::comma(round(totalSalary, 0)), "<br>Avg signing bonus: $", scales::comma(round(signingBonus, 0)), "<br>Sample size: ", n, " players"),
              color = I("darkgreen")) %>%
  add_markers(x = ~totalSalary + signingBonus,
              y = ~Position,
              text = ~paste0("Total money contract year 1: $", scales::comma(round(totalSalary + signingBonus, 0))),
              color = I("green")) %>% 
  layout(title = "Average contract and signing bonus by position",
         xaxis = list(title = "Money ($)")) %>% 
  hide_legend()

Part c – Visualize data

Create a mimimum of one additional interactive plot using plot_ly() (not ggplotly()). It can be as simple or complex as you would like. Create additional plots and find a narrative if you wish!